Max Bolingbroke <[email protected]> writes:
> It does work on HelloWorld.hs, but that program is not very
> interesting for me as it doesn't make use of any heterogeneous
> metaprogramming.

Then please try tutorial.hs using **no flags** on the command line (the
comment at the top will turn on -XModalTypes).  That file contains
plenty of code-brackets and escapes -- it undoubtedly uses the HetMet
extensions!

Type inference and checking will work exactly as advertised, but the
code you get out of the compiler will call Prelude.error.  Unfortunately
that will continue to be the case until the base library git repo is
fixed :( and I can push changes again.


> I did not realise that "flatten" was only going to be added once you
> fixed -fcoqpass. It certainly doesn't say anything like that on your
> page.

Well, there's a big red banner across the top of the page (!) explaining
that I can't push the necessary changes to fix that until the git mirror
of the base repository is updated.  If things get really, truly,
desparate I will resort to tarballs, but we're not there yet... :)


> Incidentally, the output of -dcoqpass won't compile without the
> non-standard trfrac and mathpartir style files installed.

> Perhaps it is worth linking to the sites where they can be obtained:
> http://www.utdallas.edu/~hamlen/projects.html and
> http://cristal.inria.fr/~remy/latex/

Thanks, I've added that to the web page and also put the URL in the
emitted LaTeX code.

I tried proof.sty (Makoto Tatsuta), prooftree.sty (Paul Taylor), and
bussproof.sty (Sam Buss).  I can't remember which was which but one
didn't cope well with huge proof trees, another had funny restrictions
on what kinds of things you could put in the judgments, and the third
one I just couldn't get working due probably to my own ineptitude.
Anyways I pretty quickly decided to stick with what I'm used to (trfrac)
and try to spend less time fiddling with latex.

  - a


_______________________________________________
Cvs-ghc mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/cvs-ghc

Reply via email to